Testing of networked software applications often relies on functionality and load testing. Functionality testing involves validation of the proper behavior of individual features or combinations of features. Load testing involves subjecting the application being tested to a sufficiently high number of transactions per unit time to validate that the application remains stable and meets performance requirements under this load. Nonetheless, conventional functionality and load testing can fail to fully exercise all pathways of the application, and may not accurately represent the character and volume of load that is presented to the application during actual operation in production.